Hi there,

There are several of possible reasons for that to happen. I would start with checking if the splitter has enough disk space to run by running vdf -h and df -h.

Generally I would recommend to deploy RP4VM 4.3.1 (released over a month ago) which features a Web-based Deployer with built-in verifications and generally streamlines the deployment flow quite substantially.

Regards,

Idan Kentor

RecoverPoint Corporate Systems Engineering

Thanks Idan,

I spent some time today also trying 4.3.1.  The web based deployer is a really nice feature.  Unfortunately it failed at exactly the same spot.  There is also plenty of disk space.  Any other thoughts would be appreciated.

Thanks,

Mike

The issue appears to be around exposing the repository volume to the vRPAs.  The virtual repository is created fine but it's not being discovered.  How is this virtual disk being presented to the vRPAs?

Thanks,

Mike

This also might be due to a timeout being reached when detecting the repository, if the RPAs are up and it's a VMFS as Rich suggested, the we would need logs to further investigate.

Regards,

Idan

Check that all vRPA are online, ensure you have iSCSI Software Adapters configured on the ESXi hosts and can you confirm that the datastore is a VMDK virtual device and the type.

Regards,

Rich Forshaw

Consultant Corporate Systems Engineer - RecoverPoint & VPLEX (EMEA)

Data Protection and Availability Solutions

EMC Europe Limited

Mobile: 44 (0) 7730 781169 44%20(0)%207730%20781169>

E-mail: richard.forshaw@emc.com

Twitter: @rw4shaw

Thanks Rich and Idan,

vRPAs are online and iSCSI is configured.  I also manually entered the vRPAs as targets to test iSCSI and Kashya LUNs were discovered.  The datastore is not VMFS however, it's NFS.  Could that be the issue?

Thanks,

Mike

We fully support NFS (as we're storage agnostic), as long as VMW would support the storage array, we would as well.

It's possible though that latency is high from that NFS DS, I would recommend a VMFS datastore to place the repository VMDK on, especially for troubleshooting purposes.

Regards,

Idan

Thanks Idan.  I can try iSCSI also, but if I wanted to troubleshoot NFS further, would following this be the best way to collect logs? RecoverPoint: How to collect logs for RecoverPoint.

Thanks for the support! 

The last time I worked with an NFS volume for the repository the was a timeout issue. This can be increased with assistance from Support or alternatively you can contact Idan or myself offline.

how can this timeout value be modified in 4.3.1? I found a KB that mentions dm.properties but DM is no longer involved.

Hi there,

There is a way to customize timeout values in the Web-based Deployer but it currently requires support involvement since the timeout changes are on the RPA side. Please contact Rich or myself directly to troubleshoot your specific issue.

Regards,

Idan Kentor

RecoverPoint Corporate Systems Engineering

Thanks Idan, how can I Contact you?

NP,

idan.kentor@emc.com

Regards,

Idan